4 citations · 10 across the 12 of their papers we have counts for
12 papers
Best-of-N through the Smoothing Lens: KL Divergence and Regret Analysis
Gholamali Aminian, Idan Shenfeld, Amir R. Asadi +2
A simple yet effective method for inference-time alignment of generative models is Best-of- (BoN), where outcomes are sampled from a reference policy, evaluated using a prox…
Improving Neutral Point-of-View Generation with Data- and Parameter-Efficient RL
Jessica Hoffmann, Christiane Ahlheim, Zac Yu +8
The paper shows that parameter-efficient reinforcement learning (PE-RL) is a highly effective training regime to improve large language models' (LLMs) ability to answer queries on…
Generalization and Robustness of the Tilted Empirical Risk
Gholamali Aminian, Amir R. Asadi, Tian Li +3
The generalization error (risk) of a supervised statistical learning algorithm quantifies its prediction ability on previously unseen data. Inspired by exponential tilting, \citet{…
Safety Alignment Should Be Made More Than Just a Few Tokens Deep
Xiangyu Qi, Ashwinee Panda, Kaifeng Lyu +5
The safety alignment of current Large Language Models (LLMs) is vulnerable. Relatively simple attacks, or even benign fine-tuning, can jailbreak aligned models. We argue that many…
Inducing Group Fairness in Prompt-Based Language Model Decisions
James Atwood, Nino Scherrer, Preethi Lahoti +3
Classifiers are used throughout industry to enforce policies, ranging from the detection of toxic content to age-appropriate content filtering. While these classifiers serve import…
Gradient-Based Language Model Red Teaming
Nevan Wichers, Carson Denison, Ahmad Beirami
Red teaming is a common strategy for identifying weaknesses in generative language models (LMs), where adversarial prompts are produced that trigger an LM to generate unsafe respon…
Multi-Group Fairness Evaluation via Conditional Value-at-Risk Testing
Lucas Monteiro Paes, Ananda Theertha Suresh, Alex Beutel +2
Machine learning (ML) models used in prediction and classification tasks may display performance disparities across population groups determined by sensitive attributes (e.g., race…
Improving Robustness via Tilted Exponential Layer: A Communication-Theoretic Perspective
Bhagyashree Puranik, Ahmad Beirami, Yao Qin +1
State-of-the-art techniques for enhancing robustness of deep networks mostly rely on empirical risk minimization with suitable data augmentation. In this paper, we propose a comple…
Improving Few-shot Generalization of Safety Classifiers via Data Augmented Parameter-Efficient Fine-Tuning
Ananth Balashankar, Xiao Ma, Aradhana Sinha +4
As large language models (LLMs) are widely adopted, new safety issues and policies emerge, to which existing safety classifiers do not generalize well. If we have only observed a f…
Improving Diversity of Demographic Representation in Large Language Models via Collective-Critiques and Self-Voting
Preethi Lahoti, Nicholas Blumm, Xiao Ma +8
A crucial challenge for generative large language models (LLMs) is diversity: when a user's prompt is under-specified, models may follow implicit assumptions while generating a res…
Controlled Decoding from Language Models
Sidharth Mudgal, Jong Lee, Harish Ganapathy +10
KL-regularized reinforcement learning (RL) is a popular alignment framework to control the language model responses towards high reward outcomes. We pose a tokenwise RL objective a…
Break it, Imitate it, Fix it: Robustness by Generating Human-Like Attacks
Aradhana Sinha, Ananth Balashankar, Ahmad Beirami +3
Real-world natural language processing systems need to be robust to human adversaries. Collecting examples of human adversaries for training is an effective but expensive solution.…